An Interview with Graeme Hammer

This is the first of an ongoing series of interviews with prominent thinkers in the area of climate risk management. Over the next year, we will be sharing their insights on how climate science and information can help meet the goals of development and adaptation. These individuals are pioneers in fields as diverse as climate science, sustainable development, economics and policy. We kick off this series with a Q&A with Graeme Hammer, a specialist in crop ecophysiology and modeling who sits on IRI’s International Scientific and Technical Advisory Committee.

Graeme Hammer is a professor at the University of Queensland’s School of Land, Crop and Forest Science. Among his many research interests are climate variability and seasonal climate prediction, using models to aid crop management, and regional-scale crop forecasting. The following interview took place during Dr. Hammer’s visit to the IRI in late 2008.

IRI: You are recognized for your work in using seasonal forecasts for agricultural management and planning, among other things. Could you describe that work?

Graeme Hammer: In the late 1980s, I started looking at climate risk issues because of their impact in agriculture in Northeast Australia. I was working with agricultural systems, both from the research and management perspectives with farmers and advisors. It was pretty clear that year-to-year climate variability was a huge factor in that system. The 1982 El Nino, for example, had a big impact in Australia. From one year to the next, we could see change in outcomes from one end of the spectrum to the other. For farmers in that system, it would totally dominate their perspective on profit and risk. Risk aversion through conservative practices was the approach adopted to ensure survival through the droughts. I wondered whether the emerging ability to predict seasonal climate, even though limited, might be useful in re-thinking this approach. We undertook some hypothetical studies with our agricultural systems models to simulate the likely value of adjusting management decisions based on an imperfect climate forecast. In order to do this, we developed a climate forecast system that looked at ENSO in a categorical way, based on phases of the Southern Oscillation Index. For each SOI phase, we compiled analogues of seasonal weather from historical data that we could use as input to our agricultural models. In this way, we could predict consequences of decision options associated with different categories of 3-6 month climate forecasts. The initial results were very promising, but even then we could foresee problems with explaining the necessarily probabilistic nature of the information.

IRI: Lets say we have a forecast where there is a greater chance of drier-than-normal conditions than there is of wetter-than-normal conditions. What kind of management decisions could the farmers take?

GH: In that situation farmers could change their intensity of management, to a less intensive water use layout, for example. Or they might change how much fertilizer they’d put on the crop or the type or variety of crop to plant. These things have a small but significant effect. After a while, we started working more closely with farmers on how they might use the information, rather than how we would use it or suggesting what they should do. It became pretty clear that their whole attitude toward risk drove what they did.

IRI: So if a forecast called for 60% likelihood of drier-than-normal conditions, what would farmers do?

GH: Some farmers might say, “I can use that information”, while others would have no use for it. A key part of the whole communication of using forecasts is about trying to make the point that we can’t tell you whether this season is going to be a good one or bad one. We’re only shifting the odds. Some people can handle that and others can’t. It’s about their capacity to understand what a climate forecast is and how it’s different than a simple weather forecast they might read in the paper. It’s also about how to get from the simple piece of forecast information to the understanding of whether it might change a decision. In the Australian context, much of the value for climate forecasts arises from making better use of seasons with good rainfall forecasts by increasing management intensity. This can often increase profitability without much effect on downside risk.

IRI: Oftentimes, we focus on the information gap that exists in developing countries–getting the information to the groups that can use it. But it also seems it’s not enough to just give them the information. How do you teach someone how to interpret a forecast and how to act on it?.

GH: We’re not giving people just a forecast. We’re giving them the piece of information on which they need to be able to assess their decision position. People making the decision need to know what the consequence of a forecast would be on an outcome if they were to do this or that. And they have to incorporate the uncertainly of the forecast in that analysis. This was one of the key points we tried to build into the development of the IRI and I’m really pleased to see it’s happening. It isn’t the climate forecast, stupid! Giving the farmer just a climate forecast is basically useless. The interaction of the forecast with the decision set at the farmer’s disposal is where the critical action resides. And this involves profit-risk trade-offs and personal perspectives about them. This basic overall principle is the same for all sectors. It has taken the IRI a while to evolve into this position. It changed [in 2005] from an IRI for Climate Prediction to an IRI for Climate and Society, and now you have a director [Stephen Zebiak] who travels the world talking about risk management and the needs of decision makers.

IRI: Does this make us unique in the world?

GH: It’s a very hard space in which to sit. The expectation is: you’re either a development person or research person. But to make [climate risk management] work you have to have those ends interwoven, and so you’ve got to have researchers out there in Central Kalimantan or wherever talking to the people and agencies on the ground – where the rubber hits the road, so to speak. And while the IRI can’t do the development it has to be engaged right through to that point in order to be effective. It is hard to do. You need people committed to making that happen. They might not then publish quite as many papers as they would otherwise, but they will probably make a bigger difference in the end.

IRI: What other challenges does the IRI face?

GH: The ‘climate change’ bubble has burst out into the public domain seriously. But there is a serious disconnect between what people can do now and projections of what the world might be like in 50-100 years’ time. There is a clear overlap with climate change and the climate variability/climate risk perspective. If you can manage climate risk, and if you understand the interaction of climate change with risk, then you are really positioning yourself to manage climate change. I think this is the perspective the IRI has to be able to project. No one should be able to say that climate change is really the big issue and you [IRI] are just piddling around on the margins. Working out what to do on the margins [e.g. inter-annual variability] is really the critical piece for being able to work out how to adapt climate change. That has to be really communicated.